Case studies

Client

A large regional food and restaurant group running finance, COGS and SKU profitability reporting on Azure.

Challenge

Finance, COGS, inventory and transfer-profit figures depended on complex ERP data and allocation logic. Issues were often found late, during review of management reports, and legacy Oracle and ODI processes made root causes slow to trace.

Approach

• Designed a reusable, metadata-driven data quality framework on Azure Databricks with configurable rules
• Defined validation and reconciliation rules for source-to-target accuracy, master data, calculations and missing data
• Built exception management and root-cause workflows across ERP data, transformations and allocations
• Re-engineered Oracle and ODI workloads as incremental ELT with reconciliation controls
• Coordinated remediation with finance, FP&A and business owners before production reporting

Client

A fast-growing online grocery delivery platform.

Challenge

The business needed fresher operational analytics from its PostgreSQL application databases and event streams, while platform costs were rising faster than usage.

Approach

• Architected an Azure Databricks lakehouse on Delta Lake with Event Hubs ingestion
• Implemented change data capture from PostgreSQL using Spark Structured Streaming
• Built 50+ Azure Data Factory pipelines and analytics datasets for BI
• Added validation, monitoring and reconciliation controls across pipelines
• Optimised clusters, storage and compute, and led a distributed team of 15+ engineers and analysts

Client

A mobile network operator in Iraq, engaged through a global management consultancy.

Challenge

Reporting accuracy and compliance depended on large, varied datasets without a consistent architecture for ingestion, validation and consumption.

Approach

• Designed a three-layer Bronze, Silver and Gold architecture for enterprise analytics
• Developed Data Factory and Databricks solution designs and integration patterns
• Led client discovery and architecture reviews, translating needs into data models
• Applied profiling, validation, governance and quality rules across datasets

Client

A consumer app offering loyalty and discount products across the region.

Challenge

Rapid growth created demand for customer, app and commercial analytics that ad hoc reporting could not meet.

Approach

• Led the data engineering and BI teams as data engineering manager and acting head of data
• Built an Azure data warehouse with batch and incremental pipelines and dimensional data marts
• Implemented big-data processing with Spark SQL, Python and MongoDB
• Standardised governance and reporting delivery in Power BI and Tableau